This application continues a multisite collaboration, initiated under DA 012845, to address critical issues in the genetic epidemiology of adolescent onset antisocial drug dependence. Addressing these issues requires sample sizes greater than a single site can reasonably attain, as well as the multidisciplinary expertise of psychiatrists, psychologists, and behavioral and molecular geneticists, that is difficult to provide at a single site. This collaboration includes longitudinal assessment of previously studied probands and siblings, and adds community controls. It will yield a total of ~800 clinical probands, together with their siblings, to assess differing developmental trajectories and clinical courses, and the role of comorbidity, early onset, and familial loading. After allowing for desistance, we anticipate a final sample of ~600 persistent cases, together with their siblings, and a matched sample of ~600 control subjects, for genetic association analyses of persistent, adolescent onset, antisocial substance dependence. The current application will use dense SNP association mapping to identify genetic loci predisposing to this pattern of behavior. The specific aims of the project are as follows: 1) We will complete the five year follow-up assessment of ~800 clinical probands, aged 19 though 23 years at follow-up, and their siblings, and ascertain a sample of matched control subjects from our existing databases together with 300 newly ascertained control subjects. The new assessments will be conducted at the San Diego and Denver sites. 2) We will assess differing developmental trajectories and clinical courses, and the role of comorbidity, early onset, and familial loading on these. The data set we are collecting would, even in the absence of a single DMA sample, represent a unique, and unsurpassed, research resource for studying the development and familial influences on adolescent antisocial drug dependence. 3) We will use Affymetrix SNP chip technology to genotype an average of 25 SNPs in each of 200 candidate genes for drug dependence vulnerability and/or conduct disorder. 4) We will conduct association analyses using the -600 persistent cases and their ~600 matched controls, and then conduct confirmatory tests of the best signals using a within-family association analysis (Laird and Lange, 2006). These analyses will confirm the significance and robustness of genetic associations with adolescent onset persistent antisocial drug dependence. 5) Through a continuation of our NIDA Genetics Consortium data sharing agreement, we will share all the core substance dependence and psychopathology phenotypic data, and DNA from lymphoblastoid cell lines established for the multisite samples. Our Affymetrix SNP chip will be made available, at cost, to any qualified researcher.